Customer satisfaction (CSAT)
Why it matters
A low score after a moment often comes before a customer cancels, so CSAT is an early warning for churn. It is also diagnostic. A drop after a release points at what broke, and a gap between two segments points at one of them being underserved.
Benchmarks vary widely by industry and by who is answering. B2B buyers often rate support harder than consumers do. The most useful comparison is therefore the same survey, asked the same way, over time.
How to apply it
- Attach the survey to a specific moment, such as a resolved ticket, a feature or the end of onboarding, not to the whole relationship.
- Send it straight after the experience, while it is fresh, and keep the timing the same so scores stay comparable.
- Ask for a reason when the score is low, or tag the reason afterwards: a bug, a slow reply, a delivery problem.
- Slice by agent, response time and customer type to find where the problem sits, not only whether the average moved.
- Close the loop. Tell customers what changed once a common complaint is fixed.
What it is
The standard CSAT question is simple: "How satisfied were you with this?" The answer is a rating, often from 1 to 5. The score is the share of people who gave one of the top two ratings. If 200 people answer and 160 pick a 4 or a 5, CSAT is 80 per cent.
CSAT measures one moment, not the whole relationship. That is what separates it from Net Promoter Score, which asks how likely someone is to recommend the company overall.
Common mistakes
- Surveying and then changing nothing. Customers feel heard and then ignored, which is worse than not asking.
- Averaging every score into one number that hides the cause.
- Comparing scores from different scales, such as 1 to 5 against 1 to 10.
